Saudi Journal of Ophthalmology (2011) 25, 353–362

King Saud University

Saudi Journal of Ophthalmology

www.saudiophthaljournal.comwww.ksu.edu.sa

www.sciencedirect.com

GLAUCOMA UPDATE

Gene therapy strategies in glaucoma and application

for steroid-induced hypertension

Teresa Borras, PhD *

Department of Ophthalmology, University of North Carolina School of Medicine, Chapel Hill, North Carolina, USA

Received 20 July 2011; revised 29 July 2011; accepted 30 July 2011

Available online 5 August 2011

*

C

C

St

E

13

El

Pe

do

KEYWORDS

Glaucoma;

Gene therapy;

Steroid-induced

hypertension

Address: Department of

arolina School of Medicine,

B 7041, 105 Mason Farm Roa

ates. Tel.: +1 919 843 0184;

-mail address: tborras@med

19-4534 ª 2011 King Saud

sevier B.V. All rights reserve

er review under responsibilit

i:10.1016/j.sjopt.2011.07.005

Production and h

Ophthalm

4109 Ne

d, Chape

fax: +1

.unc.edu

Universit

d.

y of King

osting by E

Abstract Gene therapy of the eye has a high potential of becoming the preferred treatment of a num-

ber of eye diseases. Because of its easy accessibility, all the tissues of the eye can be reached and genet-

ically manipulated with nowadays standard gene delivery technologies. Gene therapy offers the

possibility to do both, correct a genetic defect by replacing the mutated or missing gene and that of

using genes as drugs. Gene drugs would be more specific and would have a longer duration of action

and less toxicity than conventional drugs. Examples of both applications are beginning to emerge.

Using gene replacement, vision has been restored in several patients of Leber congenital amaurosis

(Maguire et al., 2009). Some gene drugs, such as siRNA, are currently in clinical trials to silence angio-

genic factors in macular degeneration (Campa and Harding, 2011). In this manuscript we first give a

short overview of the basics of gene therapy in the eye and then review the ongoing preclinical studies

in our laboratory for the gene-drug treatment of steroid-induced ocular hypertension.ª 2011 King Saud University. Production and hosting by Elsevier B.V. All rights reserved.

ology, University of North

uroscience Research Building

l Hill, NC 27599-7041, United

919 843 0749.

y. Production and hosting by

Saud University.

lsevier

354 T. Borras

Contents

1. Introduction and basis of gene therapy for glaucoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3542. Viral vectors and siRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354

2.1. Viral vectors to deliver genetic material to glaucoma-relevant tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3542.2. Silencing genes with short interfering RNA (siRNA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356

3. Therapeutic gene targets and regulatory elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357

3.1. Identification of genes that respond to glaucomatous insults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357

3.2. Insertion of promoter elements to specify the site and extent of gene expression . . . . . . . . . . . . . . . . . . . . . . . 357

4. Application of gene therapy to treat steroid glaucoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358

4.1. Selection of matrix metallopeptidase 1 (MMP1) to degrade extracellular matrix (ECM) in the trabecular

meshwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3584.2. Engineering of an inducible gene therapy vector to facilitate aqueous humor flow . . . . . . . . . . . . . . . . . . . . . 3594.3. Sheep model of steroid-induced ocular hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360

4.4. Lowering IOP after glucocorticoid administration with an inducible viral vector . . . . . . . . . . . . . . . . . . . . . . 3604.5. Prevention of steroid-induced elevated pressure by pre-injection with an inducible viral vector. . . . . . . . . . . . . 361

5. Conclusion and future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362

1. Introduction and basis of gene therapy for glaucoma

Glaucoma is an optic neuropathy caused by the degenerationof the retinal ganglion cells (RGCs). Glaucoma occurs morefrequently in the older population and it is the second cause

of worldwide blindness (Quigley, 1996). Blindness caused byglaucoma is irreversible. Currently, the only treatment avail-able for glaucoma is the administration of daily drops or sur-

gery. However, the use of daily drops is practically difficultand quite cumbersome for an aging individual leading to ahigh degree of non compliance. The search for an alternativetreatment where the use of the medication could be reduced

to once or twice a year is highly desirable.Using genes as drugs could bring us such possibility. Genes,

or fragments of genetic material, can be delivered inside a given

cell and allowed to make their encoded protein for long periodsof time. Because the sequence of the human genome is now com-plete, theDNAof every single enzyme or protein can be isolated,

cloned and amplified for an unlimited number of times. Properengineering of themolecule and insertion of regulatory elementsin a gene clone can theoretically control its expression and mod-ulate the targeting and abundance of the wanted product.

Althoughmultiple limiting factors and interferences do still existand detract from a perfect outcome, advances in treatment of alldiseases by gene therapy have been staggering.

In glaucoma, the major risk factor is elevated intraocularpressure (IOP). Elevated IOP is the result of an increased resis-tance of the trabecular meshwork to the flow of the aqueous

humor as it exits the eye. The trabecular meshwork is a sophis-ticated spongiform tissue formed by a loose structure of cellsand extracellular matrix (ECM) organized in characteristic

architecture. The tissue is bordered by a one-cell layer whichforms the inner wall of the Schlemm’s canal. Although everycell layer of the trabecular meshwork could be causing dys-function, it is well established that the increased resistance in

glaucoma is due in great part to an accumulation of ECMand the formation of sheath-derived plaque materials (SD-plaques) (Lutjen-Drecoll, 1973; Lutjen-Drecoll et al., 1986).

For the gene therapy treatment of glaucoma we could tar-

get either of the two main tissues: the trabecular meshworkand the RGC cells. The goal of targeting the trabecular mesh-work will be to use genes to lower IOP while the goal of target-ing RGC will be that of using genes to protect the cells from

apoptosis and death (neuroprotection) (Fig. 1).The correction of a genetic gene defect in glaucoma would

also be possible. However, this approach would most likely be

limited to single severe cases involving young patients. Themutation would have to be well characterized, and the genereplacement would have first to be conducted in an animal

model exhibiting the same gene mutation and phenotype thanthat of the patient.

2. Viral vectors and siRNA

2.1. Viral vectors to deliver genetic material to glaucoma-relevant tissues

There are four gene therapy tools being studied for genetherapy of glaucoma, three common viral vectors plus anshort interfering RNA (siRNA) (Fig. 2). The viral vectors

are adenoviruses (Ad), adeno-associated viruses (AAV) andLentiviruses. All of them have been stripped off their ownpathogenic genetic material and engineered to carry potential

therapeutic gene expression cassettes. The cassettes containthe gene to be delivered (transgene), a promoter, and possiblygene regulatory sequences.

Ads are double-stranded DNA vectors, have high tropismand efficiency for the trabecular meshwork; and have highcapacity in their genome to incorporate relative large genes.In vivo, they induce an inflammatory response at high concen-

trations and their duration of expression is short, between 1 to3 weeks in most animals (Borras et al., 2001; Kee et al., 2001).Though probably not the final tool for gene therapy, Ads are

of great use for trabecular meshwork pre-clinical studies. Stud-ies in our laboratory have investigated Ads carrying potential

Figure 2 Gene therapy tools for the trabecular meshwork. Top three diagrams: genome of the three most common viral vectors being

studied for gene delivery to the glaucoma targeting tissues. Bottom: diagram of the mechanism of silencing a gene by short interfering

RNA (siRNA); small molecules of siRNA bind to the RISC protein and are carried to the target RNA which gets degraded by the RISC

nuclease. As a consequence, the gene is silenced.

Figure 1 Target tissues for glaucomagene drugs.Left: trabecularmeshwork target; top:diagramof ameridional view section of the anterior

segment of a human eye showing the routes of aqueous humorflow; bottom: lightmicroscopy photographof ameridional section of a perfused

postmortem human anterior segment at the angle, showing the different cellular regions of the trabecular meshwork and the Schlemm’s canal

(SC). Section stainedwith toluidine blue.Right: retinal ganglion cell (RGC) layer target; top: diagramof ameridional section of the human eye

showing the delivery direction to theRGC; bottom: lightmicroscopy photograph of the human retina showing the different cell layers and and

emphasizing the RGC and nerve fiber layer which are those affected by glaucoma. Section stained with eosin and hematoxylin.

Gene therapy strategies in glaucoma and application for steroid-induced hypertension 355

candidate genes such as dominant negative RhoA, caldesmonand matrix metallopeptidase 1 (MMP1). These genes werefound to decrease trabecular meshwork outflow resistance inorgan culture (Gabelt et al., 2006; Vittitow et al., 2002) and

to reduce IOP in living animals (Gerometta et al., 2010; Spigaand Borras, 2010) (see below).

AAV are single-stranded DNA vectors, have high tropismfor the retina, and have the best gene therapy safety record

356 T. Borras

to date. Once into the cells, AAV can express the transgene forup to five years after a single dose (Rivera et al., 2005). AnAAV2 vector carrying the RPE65 gene is currently being used

in clinical trials to reverse congenital blindness caused by a ret-inal pigmented epithelium cell defect (Stein et al., 2011).Numerous examples of AAV vectors carrying neuroprotective

genes (such as ciliary neurotrophic factor, CNTF or brain-de-rived neurotrophic factor, BDNF) and injected intravitreallyhave shown to protect RGC from rat hypertensive models

(Hellstrom and Harvey, 2011). Very recently, an exciting studyusing systemic delivery of an AAV vector carrying the pig-ment-derived epithelium factor (PEDF) transgene was ableto protect RGC death in the glaucomatous DBA/2J mouse

(Sullivan et al., 2011).AAV viral vectors though, do not transduce the trabecu-

lar meshwork due to their inability to form a double

stranded DNA upon entering the cell. However, becauseof their high desirability for long-term and safe gene therapyprofile, a second generation AAV vector has been developed,

the self-complementary AAV (scAAV) which overrides thelimiting step. A single intracameral dose of scAAV.GFP inliving monkeys resulted in positive gene delivery to the tra-

becular meshwork with early onset and lasting for at leasttwo years (Buie et al., 2010) (Fig. 3). These encouragingfindings are setting the way for safe gene targeting of thetrabecular meshwork.

Lentiviruses are single stranded RNA vectors derived fromhuman or simian immunodeficiency viruses. They integrateinto the host genome and as a result, retain the ability to ex-

press the transgene for a long time. Lentiviruses transduceboth the trabecular meshwork and the RGCs, and these vec-tors carrying genes for cyclooxigenase-2 (COX-2) and pig-

ment-derived epithelium factor (PEDF) lowered IOP in cats

Figure 3 Expression of reporter transgene after single intracameral inj

fluorescent protein (GFP) to the anterior segment of the living rat (left

Images of rat frozen meridional sections at different time points were t

stable and long-term expression was monitored noninvasively with gon

gene transfer, represented by the presence of green cells occurred in liv

months in rats and for over two years in monkeys. Contralateral eyes in

and protected death of RGC in rats, respectively (Barrazaet al., 2010; Miyazaki et al., 2011).

2.2. Silencing genes with short interfering RNA (siRNA)

Short interfering siRNAs are double stranded RNA molecules21–24 nucleotides long. Inside the cells siRNA interacts with a

nuclease-containing multiprotein complex called RISC (RNA-induced silencing complex). After binding to RISC, the siRNAunwinds and pairs with its target mRNA; the RISC nuclease

cleaves the mRNA at the target site and this cleavage precipi-tates a full degradation of the mRNA molecule which now isunable to translate and produce the protein. siRNAs are thus

a powerful tool to silence unwanted genes and/or modulateexpression of downstream cascades. In initial experiments,the perfusion of post-mortem human anterior segments witha fluorescent Cy3 labeled naked siRNA resulted in the en-

trance of the molecule into the trabecular meshwork tissue.Subsequently, perfusion of the siRNA for the glucocorticoidreceptor (GR) did degrade its own RNA. Furthermore, in

the presence of dexamethasone (DEX), the GR siRNA modu-lated also the expression of two DEX-induced genes (Myocilinand Angiopoietin-like7). That is, silencing the receptor did

control the response of other genes induced by the glucocorti-coid (Comes and Borras, 2007). Continuation of these studiesin living animals is showing positive delivery of siRNA byintracameral injection to the trabecular meshwork of rats

(Lawrence et al., 2011). siRNA strategies are being developedby several pharmaceutical companies. For glaucoma, a clinicaltrial study presented at the recent ARVO meeting, showed

reduction of IOP upon administration of eye drops containinga naked siRNA to the b2 adrenergic receptors (Ruz et al.,2011). Studies coupling siRNA to different polymers and

ection of scAAV.GFP. Delivery of the reporter gene encoding green

) and living monkey (right) after single injection of the viral vector.

aken under a fluorescent microscope (left). In the monkeys (right),

iophotography; images were taken with a fundus camera: positive

ing animals. The expression of the transgene lasted for about three

jected with phosphate-buffered saline (PBS) were negative.

Gene therapy strategies in glaucoma and application for steroid-induced hypertension 357

forming nanoparticles are part of a big effort to increase effi-ciency of delivery of molecules (Huang and Liu, 2011).

3. Therapeutic gene targets and regulatory elements

In parallel to optimizing delivery vehicles that will carry genesinto the cells, our investigations and those of others are being di-

rected toward the identification of gene targets which will lowerIOP by modifying the properties of the trabecular meshworkcells. Because the function of any tissue is given by the expres-

sion of its genes, we first looked at the most abundant genes ex-pressed in the trabecularmeshwork under normal, physiologicalconditions. To understand which genes would be relevant in the

cause and response of the cells to glaucoma, we performed sim-ilar studies subjecting the trabecular meshwork tissue to glauco-matous insults, such as elevated pressure and glucocorticoids.

For a human model, we used the perfused anterior segment per-fusion where tissues from donor eyes, up to 35 h post-mortemcan be revived in organ culture.Most important in such amodelis the fact that the gene response of a treated eye (OD) can be

compared to that of paired contralateral eye (OS) which hasan identical genetic background. The results are then not con-founded by genetic differences among individuals (Fig. 4).

3.1. Identification of genes that respond to glaucomatous insults

Microarray gene profiles performed with the RNA extracted

from trabecular meshwork of different individuals subjectedto an elevated IOP insult (Fig. 5) reveal a set of candidate genesthat we termed the ‘‘IOPmolecular biomarkers of the trabecularmeshwork’’ (Fig. 5) (Comes and Borras, 2009). This set of genes

has been cross-checked with other gene sets altered by treat-ments with TGFb2, DEX and with those which are present only

Figure 4 Human anterior segment perfusion model. Top: preparation

30–40 h postmortem are bisected at the equator and have their lens, iris

custom-made 2-cannula polycarbonate dish and secured by an open ring

maintained inside a CO2 incubator and perfused through one of the ca

second cannula is connected to a pressure transducer to monitor IOP. G

loop, that is intercalated between the syringes and organ culture cham

in the trabecular meshwork from glaucoma patients (Borras,2008). The result of the cross-cross-check yielded a gene list rep-resenting a ‘‘molecular signature of glaucoma’’ (Borras, 2008).

An interesting finding from these studies is that genes alteredby high IOP could be divided into two subgroups, individualand general responders (Comes and Borras, 2009). This ability

of some genes to respond differently to pressure in some individ-uals provides the firstmolecular explanation to the different IOPoutcomes observed in the clinic. Interestingly, the 10 selected

genes identified in Fig. 5 are representative of several cellularmechanisms that could affect outflow facility. Angiopoietin-like7 and matrix metallopeptidases 1 and 12 (MMP1 and MMP12)would affect the remodeling of the ECM; lysyl oxidase and fib-

ulin 5 would affect the collagen-elastin network; a-B crystallinand myocilin would be involved in stress response; matrix Glawould affect the calcification state of the cell; podoplanin and

chemokine CXCL2 would affect unknown pathways involvedwith the lymphatic and immune systems.

3.2. Insertion of promoter elements to specify the site and extentof gene expression

Two very important parameters to consider during the devel-

opment of a gene therapy regimen are tissue targeting and reg-ulation of the expression of the gene. To avoid unwanted,secondary effects due to gene-drug expression in surroundingtissues, tissue-specific elements need to be introduced in the

gene promoter that would direct its expression only to thewanted tissue. In addition, because gene-drugs could have avery long duration of action (years), specific regulatory

elements need to be introduced in its promoter to be able toturn them on and off at will. Ideally, a gene should be turnedon ‘‘automatically’’, in the presence of the glaucomatous in-

of the postmortem human eyes for perfusion. Donor human eyes,

and vitreous removed. The anterior segment is then mounted to a

. Bottom: diagram of the perfusion system. The cultured chamber is

nnulas at constant flow using a microinfusion syringe pump. The

ene drugs are injected through a HPLC pump equipped with a 20-llbers. Pumps are controlled by a custom-made computer program.

Figure 5 Identification of pressure responder genes. (A) Diagram of the anterior segment perfusion model during an elevated IOP

experiment. To raise IOP, the flow rate of one anterior segment is increased to achieve a DP of �35 mmHg. The flow rate of the

contralateral eye remains at baseline. (B) Representative profile of a human eye pair perfused at elevated pressure for three days. IOP is

continuously monitored by a pressure transducer and plotted with values obtained every 30 min. (C) Ten selected pressure responder genes

obtained from microarray profiles of pressured insulted trabecular meshworks (Affymetrix).

358 T. Borras

sult, and turned off when the insult is no longer present. In

some cases, those inducible elements are very well defined inthe literature as in the case of glucocorticoid response elements(GRE, see below) (Aranda and Pascual, 2001). Other elements

needed for our glaucoma application need to be identified. Anumber of such vectors are currently in the pipeline. A vectortargeting the trabecular meshwork is carrying the promoter of

a B-crystallin driving what is called a reporter gene for easydetection in histological sections. The strategy of this tissue-targeting vector is based on a comprehensive review of theexpression profile of genes which are not expressed in tissues

facing the anterior chamber, and in the logistics that Ad vec-tors do not penetrate the lens capsule (where aB-crystallin ishighly expressed). A vector responding to glucocorticoids is

carrying the glucocorticoid response element (GRE) in frontof a basal promoter and driving the MMP1 potential therapeu-tic gene. Such a vector has been fully characterized and proved

to be specific in an animal model of hypertension (see section4). An additional vector where the expression of the gene willdepend of whether elevated IOP is present, is also under devel-opment. Since no IOP responding elements are yet defined, the

beginning strategy has been to use the promoter of a generalpressure responder gene. The first vectors contain the promot-ers of matrix Gla and angiopoietin-like 7 driving a reporter

gene (secreted alkaline phosphatase) whose expression can bemeasured biochemically along the time of high IOP exposure.

4. Application of gene therapy to treat steroid glaucoma

There is a very good rationale as to why the development of agene therapy regimen for steroid glaucoma could result in an

attractive and efficient way to manage this type of ocularhypertension. Two percent of the general population receivesglucocorticoid treatments. Thirty to 40% of the steroid-treated

patients develop elevated IOP (Armaly and Becker, 1965;

Johnson, 1997). At the cellular level, it is well established that

steroid treatment leads to an increase of ECM material whichmimics that of primary angle glaucoma (Johnson et al., 1997).This fact implies that several ECM targets could be used to

counteract the excessive build up. At the molecular biology le-vel, the promoter sequences that respond to glucocorticoidsand turn on gene expression have been very well defined in sev-

eral other cells. Putting all this information together led ourlaboratory to develop a strategy where GRE elements wouldbe inserted in front of an ECM degradation gene. To then in-sert such a cassette into a gene therapy vector and to inject the

vector intracamerally into the eye of an animal model of ocularhypertension. The intent of the project was to induce a reorga-nization of the ECM only in the presence of the glucocorticoid.

The final goal was to reduce and prevent the steroid-inducedelevated IOP.

4.1. Selection of matrix metallopeptidase 1 (MMP1) to degradeextracellular matrix (ECM) in the trabecular meshwork

Matrix metallopetidases (MMPs) comprise a family of zinc-

binding proteases known to play a key role in the turnoverof the ECM of the trabecular meshwork (Keller et al., 2009).Previous studies on MMPs have shown that these enzymesare able to increase aqueous humor outflow in perfused organ

cultures (Bradley et al., 1998). A member of this family isMMP1, an interstitial collagenase which breaks down colla-gens type I, II and III. Collagen type I is a main component

of the trabecular meshwork’s ECM scaffold. It constitutesthe central core of the trabecular meshwork beams. It wasknown that treatment of trabecular meshwork with DEX in

organ culture conditions up-regulates collagen type I anddown-regulates MMP1 (Rozsa et al., 2006; Zhou et al.,1998). Further, treatment of human trabecular meshwork cells(HTM) with triamcinolone and prednisolone also down-

Gene therapy strategies in glaucoma and application for steroid-induced hypertension 359

regulates MMP1 (Spiga and Borras, 2010). Therefore, MMP1was selected as the gene target to investigate whether itsexpression would counteract the steroid effect in living ani-

mals. Overexpressing MMP1 under glucocorticoid conditionswould serve not only to counteracting its down regulation,but also would help reduce the ECM accumulation.

4.2. Engineering of an inducible gene therapy vector to facilitate

aqueous humor flow

An adenoviral vector containing an inducible MMP1 cDNAwas generated using recombinant DNA techniques. The fullcoding MMP1 cDNA (1410 nt), able to transcribe and trans-

late a pre-active protein, was amplified from laboratoryHTM cells overexpressing MMP1. The basic promoter andGRE sequences, able to induce a downstream gene, were ex-tracted from a commercial vector. Fusion of these two DNA

fragments formed the expression cassette which will be incor-porated into the viral vector Adh.GRE.MMP1. A second vec-tor producing a mutated MMP1 protein was also generated

and used as negative control. This mutated MMP1 contained

Figure 6 Generation and assays of inducible gene therapy vectors

representation of glucocorticoid-inducible viral vectors expressing reco

MMP1 only after DEX binding to the GRE sequences in the promoter.

cells. Cells were infected with wild-type and mutant viral vector, treated

MMP1 RNA assayed by TaqMan PCR showed a very significant incre

As well, levels of MMP1 protein assayed by western blots were highl

activity of secreted recombinant MMP1 in HTM cells. Cells were inf

harvested. Aliquots of the cell media were activated and incubated w

stained with Comassie blue to assay intact or degraded collagen. Only a

wild-type vector were able to break down collagen (lane 2).

one nt substitution at the catalytic binding site which leadsto improper folding and destroys its catalytic activity(Adh.GRE.mutMMP1) (Fig. 6A) (Spiga and Borras, 2010).

Response of the potential vectors to glucocorticoids wasextensively characterized in primary cell and organ culturesconditions. As expected, infection of the cultures with both

vectors induced high levels of MMP1 mRNA and protein(Fig. 6B). However, only the protein produced by the wild-type MMP1 had collagenase activity and was able to degrade

collagen. This important property was tested in several as-says. In a classical assay, purified rat collagen was incubatedwith the activated conditioned media of cells infected withMMP1 vectors and ran in polyacrylamide gels. Rat collagen

was seen degraded only in the gel lane which had beenloaded with the activated wild-type protein (Fig. 6C). Amore sophisticated technology, state of the art assay, the

fluorescence resonance energy transfer (FRET), utilized anMMP substrate peptide labeled with a fluorophore and aquencher. Cleavage of the peptide with MMP1, which re-

leases the fluorophore and is read in a fluorophotometer,was observed only in the incubation of the wild-type-infected

for the treatment of steroid-induced hypertension. (A) Schematic

mbinant MMP1. Wild-type (active) and mutant vectors expressing

(B) DEX-induced overproduction of recombinant MMP1 in HTM

with 0.1 lM DEX and harvested for RNA and protein. Levels of

ase in the DEX treated cells compared to untreated controls (left).

y increased in the treated cells (right). (C) Enzymatic collagenase

ected and treated with DEX as in B and conditioned media was

ith rat tail native collagen for 2 h. Samples were run on gels and

ctivated samples obtained from DEX-treated cells infected with the

360 T. Borras

conditioned media with the FRET peptide. Lastly, by immu-nohistochemistry, double staining of the human perfusedtrabecular meshwork with MMP1 and collagen type I anti-

bodies revealed degradation of collagen in the areas wherelevels of MMP1 were higher (Spiga and Borras, 2010).

Another important property of the Adh.GRE.MMP1 vec-

tor was the fact that overexpression of the MMP1 transgenecould be synchronized with the administration of DEX. Uponinfection of HTM cells with the vector, MMP1 expression was

consecutively on and off coinciding with the treatment andwithdrawal of the glucocorticoid. For a gene therapy drugapplication, this characteristic is of upmost importance. Deg-radation of the ECM in steroid-induced hypertension will oc-

cur only in the presence of the steroid.Altogether the generated vector Adh.GRE.MMP1 seemed

to contain all characteristics required for its application to

an steroid-induced in vivo model.

4.3. Sheep model of steroid-induced ocular hypertension

The laboratories of Drs. Oscar Candia (Mount Sinai School ofMedicine) and Rosana Gerometta (Universidad Nacional delNordeste, Corrientes Argentina) have recently developed a

model of glucocorticoid induced ocular hypertension in Cor-riedale sheep (Ovis aries) (Gerometta et al., 2009). Topicalapplication of two drops of 0.5% prednisolone acetate (Ultra-corteno, Novartis) 3· daily into one eye of the sheep results in

a �2.5· IOP increase within 1–2 weeks. IOP values, read witha Perkins applanation tonometer, are between 9–11 mmHg at

Figure 7 Adh.GRE.MMP1 injection lowers steroid-induced hypert

values from two sheep: (A) sheep injected in one eye with the active

mutant Adh.GRE.mutMMP1 (B). Contralateral eyes also treated with

IOP measurements. Only the eye injected with the active gene therapy

steroids.

baseline and raise to 25–28 mmHg. Pressures of the contralat-eral eye, receiving artificial tears, remain at baseline levels. Incontrast to humans, this elevation of IOP occurs in 100% of

the glucocorticoid treated animals. The elevated IOP is main-tained during a 4 week application of the prednisolone drops,persists after discontinuation of the treatment, and returns to

baseline levels over the course of one to three weeks(Gerometta et al., 2009).

An elevation of IOP is also obtained by a single sub-Tenon

injection of triamcinolone. One milliliter injection of triamcin-olone acetonide (40 mg/ml, Bristol-Myers Squibb) is adminis-tered under topical anesthesia using a 30G needle. Theinjection is performed to create a juxta-sclera depot rather that

an intra Tenon injection (Gerometta et al., 2010). Pressures ofthe injected eye are increased �2· at day 4 post-injection andremain high for about 2–3 weeks.

The docile nature of these animals together with their 100%response to the steroid makes the sheep model ideal for studiesof gene therapy treatment of steroid-induced hypertension.

4.4. Lowering IOP after glucocorticoid administration with an

inducible viral vector

In a study involving a total of six sheep, baseline pressureswere taken for one week and prednisolone was administereddaily afterward in both eyes. At four days post-prednisolonetreatment, when the pressures had risen to �2· baseline levels,

one eye of each sheep received an intracameral injection of 1 of3 adenoviral vectors while the contralateral eyes remained

ension in sheep. Representative corticosteroid regimen and IOP

Adh.GRE.MMP1. (B) Sheep injected in one eye with the inactive

corticosteroid, remained uninjected. (C) Representative plots of the

vector experimented a decrease from the elevated IOP induced by

Figure 8 Prevention of steroid-induced hypertension. Pre-injection of the active gene therapy vector Adh.GRE.MMP1 prevents the

elevated IOP induced by the corticosteroid. Contralateral uninjected eye received two corticosteroid treatments and induced elevated IOP.

Gene therapy strategies in glaucoma and application for steroid-induced hypertension 361

uninjected. One vector carried the wild-type MMP1 (Ad-h.GRE.MMP1) (Fig. 7A), the second vector carried the mu-

tated form of the protein lacking the active catalytic site(Adh.GRE.mutMMP1) (Fig. 7B), and the third vector was‘‘empty’’, that is, carrying no transgene (Ad.Null). The contra-

lateral eyes were left uninjected. The IOP stayed elevated in alleyes that were either uninjected or injected with the controlvectors for the duration of the prednisolone instillation How-

ever, in eyes injected with the active MMP1 the elevated IOPreturned to the lower baseline values two days after the viralinjection (Gerometta et al., 2010). Low pressures of the wild-

type injected eyes persisted for approximately 10–15 days long-er, which correspond with the published data of the durationof the expression of adenoviruses in living animals (Fig. 7)(Borras et al., 2001; Kee et al., 2001).

An additional four sheep received sub-Tenon injections oftriamcinolone in both eyes. Four days later, IOP increasedfrom 11.3 ± 0.3 to 22. ± 0.8 mmHg in one eye and, from

9.7 ± 0.2 to 22.1 ± 0.2 mmHg in its contralateral. One eyeof each sheep was then injected with the wild-type Ad-h.GRE.MMP1 while the paired eye was injected with the vec-

tor carrying the mutated form of the protein (Adh.GRE.mutMMP1). Two to three days after viral injections, IOPwas significantly reduced in the eyes injected with the activeMMP1 (11.6 ± 0.4) but remained elevated in the eyes injected

with vector carrying the mutant protein which lacks the cata-lytic site (22.1 ± 0.2 mmHg). However, in contrast with theprednisolone experiments, where the reduction was maintained

for almost two weeks, the reduction observed with the wild-type vector lasted just for three to four days (Geromettaet al., 2010). The causes of the short reduction after triamcin-

olone injection are not known and need further investigation.Perhaps, variability in the formation of the depot and releaseof the steroid plays a role in overriding the effect of the vector.

None of the sheep eyes injected with any of the adenoviralvectors showed adverse clinical signs. There were no signs of

hyperemia or inflammation, and the corneas remained clear.This first case of lowering elevated IOP by a gene therapy vec-

tor in a large animal provides the initial steps for the develop-ment of a new avenue of treatment for ocular hypertensioninduced by steroids.

4.5. Prevention of steroid-induced elevated pressure by pre-

injection with an inducible viral vector

In addition to lowering the steroid-induced IOP, the adenovi-ral vector which delivers the active MMP1 is also able to pre-vent the steroid-induced IOP elevation. In two independent

experiments, Adh.GRE.MMP1 was injected at day 0 and pred-nisolone instillation started two or three days after injection.Pressure elevation did not occur until 12 to 15 days post-viralinjections, in concordance with the duration of expression of

the transgene in an adenoviral vector (Fig. 8).Similarly, the same Adh.GRE.MMP1 vector was able to

avoid the doubling pressure increase induced by the triamcino-

lone. At day 0, with baselines pressures of 9.4 ± 0.1 mmHg, theeyes of two sheep were injected with the active viral vectors fol-lowed by triamcinolone injections one day later. Pressures re-

mained close to baseline levels for four days, reached a mid-value of 12.9 ± 0.5 mmHg at five days and attained the ex-pected triamcinolone doubling response of 21.3 ± 0.3 mmHgat 11 days. These findings indicate that during the time the virus

is producing the therapeutic protein (10–15 days), there is acounteractive, protective action that impedes the steroid to exertits full potential.

5. Conclusion and future directions

During the past few years, there have been important advances

in the use of gene therapy for the treatment of eye diseases.Without a doubt, the most important success has been the res-

362 T. Borras

toration of vision by the replacement of the RPE65 gene to pa-tients of Leber congenital amaurosis (Stein et al., 2011). Inaddition of demonstrating efficacy by delivering the wild-type

gene, these clinical trials have served to validate the safety ofthe AAV viral vector for the use of gene therapy in the humaneye. The importance of this safety issue cannot be underesti-

mated. Regarding glaucoma, we have made considerable pre-clinical advances in all four relevant fronts: potentialtherapeutic genes with regulatory elements, safe and efficient

delivery vectors, animal models, and significant improvementsin animal evaluation technology (from physiology to imaging).There are now a considerable number of genes with neuropro-tective properties under study in several laboratories. These

genes are being tested in AAV vectors, which transduceRGC cells very efficiently. There are also, as we saw above,potentially therapeutic genes controlled by regulatory elements

that could target the trabecular meshwork and lower elevatedIOP. We do have a second generation AAV vector to trans-duce the outflow tissue and we are beginning to have good ani-

mal models to test their efficiency. An important body of workis ahead and we are not yet ready. However, the generaloutlook is that the gene therapy approach for glaucoma looks

much more feasible now than what it did five years ago. Wecan only envision that the next five years will be as productiveand will take us to the trial of the first gene therapy drug forglaucoma in the clinical setting.

Acknowledgments

These studies were supported by National Institutes of Healthgrants EY11906 and EY13126 (T.B.) plus a Research to Pre-

vent Blindness (RPB) unrestricted grant to the UNC depart-ment of ophthalmology.

References

Aranda, A., Pascual, A., 2001. Nuclear hormone receptors and gene

expression. Physiol. Rev. 81, 1269–1304.

Armaly, M.F., Becker, B., 1965. Intraocular pressure response to

topical corticosteroids. Fed. Proc. 24, 1274–1278.

Barraza, R.A. et al., 2010. Prostaglandin pathway gene therapy for

sustained reduction of intraocular pressure. Mol. Ther. 18, 491–501.

Borras, T., 2008. What is functional genomics teaching us about

intraocular pressure regulation and glaucoma. In: Civan, M.M.

(Ed.), The Eye’s Aqueous Humor, second ed. Elsevier, San Diego,

pp. 323–377.

Borras, T. et al., 2001. Non-invasive observation of repeated adeno-

viral GFP gene delivery to the anterior segment of the monkey eye

in vivo. J. Gene Med. 3, 437–449.

Bradley, J.M. et al., 1998. Effect of matrix metalloproteinases activity

on outflow in perfused human organ culture. Invest. Ophthalmol.

Vis. Sci. 39, 2649–2658.

Buie, L.K. et al., 2010. Self-complementary AAV virus (scAAV) safe

and long-term gene transfer in the trabecular meshwork of living

rats and monkeys. Invest. Ophthalmol. Vis. Sci. 51, 236–248.

Campa, C., Harding, S.P., 2011. Anti-VEGF compounds in the

treatment of neovascular age related macular degeneration. Curr.

Drug Targets 12, 173–181.

Comes, N., Borras, T., 2007. Functional delivery of synthetic naked

siRNA to the human trabecular meshwork in perfused organ

cultures. Mol. Vis. 13, 1363–1374.

Comes, N., Borras, T., 2009. Individual molecular response to elevated

intraocular pressure in perfused postmortem human eyes. Physiol.

Genomics 38, 205–225.

Gabelt, B.T. et al., 2006. Caldesmon transgene expression disrupts

focal adhesions in HTM cells and increases outflow facility in

organ-cultured human and monkey anterior segments. Exp. Eye

Res. 82, 935–944.

Gerometta, R. et al., 2009. Steroid-induced ocular hypertension in

normal sheep. Invest. Ophthalmol. Vis. Sci. 50, 669–673.

Gerometta, R. et al., 2010. Treatment of sheep steroid-induced ocular

hypertension with a glucocorticoid-inducible MMP1 gene therapy

virus. Invest. Ophthalmol. Vis. Sci. 51, 3042–3048.

Hellstrom, M., Harvey, A.R., 2011. Retinal ganglion cell gene therapy

and visual system repair. Curr. Gene Ther. 11, 116–131.

Huang, L., Liu, Y., 2011. In vivo delivery of RNAi with lipid-based

nanoparticles. Annu. Rev. Biomed. Eng. 13, 507–530.

Johnson, D.H., 1997. Corticosteroid glaucoma. In: Epstein, D.L.,

Allingham, R.R., Schuman, J.S. (Eds.), Chandler and Grant’s

Glaucoma. Williams & Wilkins, Baltimore, Maryland, USA, pp.

404–411.

Johnson, D. et al., 1997. Ultrastructural changes in the trabecular

meshwork of human eyes treated with corticosteroids. Arch.

Ophthalmol. 115, 375–383.

Kee, C. et al., 2001. Stromelysin gene transfer into cultured human

trabecular cells and rat trabecular meshwork in vivo. Invest.

Ophthalmol. Vis. Sci. 42, 2856–2860.

Keller, K.E. et al., 2009. Extracellular matrix turnover and outflow

resistance. Exp. Eye Res. 88, 676–682.

Lutjen-Drecoll, E., 1973. Structural factors influencing outflow facility

and its changeability under drugs. A study in Macaca arctoides.

Invest. Ophthalmol. 12, 280–294.

Lutjen-Drecoll, E. et al., 1986. Quantitative analysis of ‘plaque

material’ in the inner- and outer wall of Schlemm’s canal in

normal- and glaucomatous eyes. Exp. Eye Res. 42, 443–455.

Maguire, A.M. et al., 2009. Age-dependent effects of RPE65 gene

therapy for Leber’s congenital amaurosis: a phase 1 dose-escalation

trial. Lancet 374, 1597–1605.

Miyazaki, M. et al., 2011. Pigment epithelium-derived factor gene

therapy targeting retinal ganglion cell injuries: neuroprotection against

loss of function in two animal models. Hum. Gene Ther. 22, 559–

565.

Quigley, H.A., 1996. Number of people with glaucoma worldwide. Br.

J. Ophthalmol. 80, 389–393.

Rivera, V.M. et al., 2005. Long-term pharmacologically regulated

expression of erythropoietin in primates following AAV-mediated

gene transfer. Blood 105, 1424–1430.

Rozsa, F.W. et al., 2006. Gene expression profile of human trabecular

meshwork cells in response to long-term dexamethasone exposure.

Mol. Vis. 12, 125–141.

Ruz, V. et al., 2011. Phase I study with a new siRNA: SYL040012.

Tolerance and effect on intraocular pressure. Invest. Ophthalmol.

Vis. Sci. (ARVO Abstract).

Lawrence, S.D. et al., 2011. Direct delivery of synthetic naked siRNA

to reabecular meshwork in living rats. Invest. Ophthalmol. Vis. Sci.

(ARVO Abstract).

Spiga, M.G., Borras, T., 2010. Development of a gene therapy virus

with a glucocorticoid-inducible MMP1 for the treatment of steroid

glaucoma. Invest. Ophthalmol. Vis. Sci. 51, 3029–3041.

Stein, L. et al., 2011. Clinical gene therapy for the treatment of

RPE65-associated Leber congenital amaurosis. Expert Opin. Biol.

Ther. 11, 429–439.

Sullivan, T.A. et al., 2011. Systemic adeno-associated virus-mediated gene

therapy preserves retinal ganglion cells and visual function in DBA/2J

glaucomatous mice. Hum. Gene Ther. [Epub ahead of print].

Vittitow, J.L. et al., 2002. Gene transfer of dominant-negative RhoA

increases outflow facility in perfused human anterior segment

cultures. Mol. Vis. 8, 32–44.

Zhou, L. et al., 1998. Glucocorticoid effects on extracellular matrix

proteins and integrins in bovine trabecular meshwork cells in

relation to glaucoma. Int. J. Mol. Med. 1, 339–346.